Using ClickHouse To Power Crypto Analytics For The Leading Hyperliquid Blockchain

Hyperliquid is rapidly becoming one of the leading layer 1 blockchains and Crypto projects. It is known for its very low latency and fees which make it attractive to traders and institutions. It is growing rapidly by all metrics including volume traded, number of trades and unique users.

Our client are using ClickHouse to deliver analytics based on data captured from the Hyperliquid network. This involves the real time ingestion of data from the blockchain and validators and the creation of real time analytics using materialised views. This data is then served to a frontend web application directly from ClickHouse.

The project initially began with performance optimisations of their existing solution. We then developed a series of materialized views to generate both vanilla trading metrics (HLOC, trending tokens, volumes) before moving to more sophisticated trading and risk management signals.

The customer were experiencing the following challenges prior to the engagement:

• Performance issues as a result of their high frequency ingest.
• Difficulty completing the complex SQL to capture detailed analytics.
• The need to use advanced features of ClickHouse to process the high data volumes with acceptable latency.
• The need to serve their ClickHouse data to their web application clients in an effective way.

We took the following approach to this project:

• Investigated the cause of the performance issue and identified root causes including contention between readers and writers. Recommended a clustered approach to separate readers and writers.
• Optimised inserts to reduce background activity in ClickHouse.
• Developed the new materialised views for real time analytics, trading and risk management signals.
• Implemented an innovative API to make it easy for web application developers to consume ClickHouse hosted data and analytics.